Competition dynamics of second-language listening

Abstract

Spoken-word recognition in a nonnative language is particularly difficult where it depends on discrimination between confusable phonemes. Four experiments here examine whether this difficulty is in part due to phantom competition from “near-words” in speech. Dutch listeners confuse English /æ/ and /ϵ/, which could lead to the sequence daf being interpreted as deaf, or lemp being interpreted as lamp. In auditory lexical decision, Dutch listeners indeed accepted such near-words as real English words more often than English listeners did. In cross-modal priming, near-words extracted from word or phrase contexts (daf from DAFfodil, lemp from eviL EMPire) induced activation of corresponding real words (deaf; lamp) for Dutch, but again not for English, listeners. Finally, by the end of untruncated carrier words containing embedded words or near-words (definite; daffodil) no activation of the real embedded forms (deaf in definite) remained for English or Dutch listeners, but activation of embedded near-words (deaf in daffodil) did still remain, for Dutch listeners only. Misinterpretation of the initial vowel here favoured the phantom competitor and disfavoured the carrier (lexically represented as containing a different vowel). Thus, near-words compete for recognition and continue competing for longer than actually embedded words; nonnative listening indeed involves phantom competition.

Keywords:

• Word recognition
• Competition
• Listening
• First language
• Second language

Acknowledgments

This research was supported by a VENI grant from the Netherlands Organisation for Scientific Research (NWO) to the first author, and by a SPINOZA award from the NWO to the second author. We would like to thank Antje Meyer of the Behavioral Brain Science Center, University of Birmingham, and Alan Garnham of the Laboratory of Experimental Psychology, University of Sussex, for giving us the opportunity to test participants in their labs. Thanks also to Nicolas Dumay and two anonymous reviewers for useful comments.

Notes

¹ We thank reviewer Nicolas Dumay for this suggestion.